Multi-Model Inference for Estimating Entrainment Survival

Historically, fairly simple statistical methods, such as Abbott’s formula (Abbott 1925), have been used to estimate survival of fish and invertebrates entrained into power plant cooling systems.  Similar to analysis of bioassay data, Abbott’s formula separates mortality due to the treatment (entrainment) from that occurring in control samples, under the assumption that the two mortality sources are independent.   More recent statistical methods (information theoretic methods combined with failure time analysis) are more useful and flexible for analysis of entrainment (or impingement) survival data.   Examples of how these methods are able to improve entrainment survival estimates, including confidence bounds around estimates, are provided.